Structured entry of narrative data

Structured narrative entry is designed for use while an interview is taking place, including a telephone interview. The interviewer transcribes the account as the person speaks.

Before entering a person, location or another supported element, the interviewer types ... A menu opens at the cursor. The interviewer selects the category, enters the details and inserts the element into the narrative, then continues transcribing.

The structured record is created at that point in the conversation. The narrative and structured data are entered together rather than being processed in a separate step after the interview.

The prototype explores:

• Transcribing an account in the order it is given.
• Categorising people, places, times and transport details without leaving the narrative.
• Creating or reusing structured records as each element is mentioned.
• Keeping the interviewer’s hands on the keyboard and the conversation in view.
• Retaining approximate language such as “around 1pm” in the structured entry.

The interaction uses guided operator entry. It does not automatically extract or interpret the interview using AI.

Location lookup can depend on a configured external service. Manual entry remains available when a lookup is unavailable or unsuitable.

The demonstration illustrates the proposed interaction. It does not establish that the feature is available in the current Sentinel release.

Illustrative walkthrough

The interviewer types .. before entering the element: choose Person, enter Alex, insert the person, then continue the transcript. Structured entry happens during the interview.